Investigation of the ATP Binding Site of Escherichia coli

Aminoimidazole ribonucleotide (AIR) synthetase (PurM) catalyzes the conversion of ... and ATP to AIR, ADP, and Pi, the fifth step in de novo purine bi...
0 downloads 0 Views 177KB Size
Biochemistry 1999, 38, 9831-9839

9831

Investigation of the ATP Binding Site of Escherichia coli Aminoimidazole Ribonucleotide Synthetase Using Affinity Labeling and Site-Directed Mutagenesis† E. J. Mueller,‡ S. Oh,‡ E. Kavalerchik,‡ T. J. Kappock,‡ E. Meyer,‡ C. Li,§ S. E. Ealick,*,§ and J. Stubbe*,‡ Departments of Chemistry and Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, and Department of Chemistry and Chemical Biology, Cornell UniVersity, Ithaca, New York 14853 ReceiVed March 18, 1999; ReVised Manuscript ReceiVed May 26, 1999

ABSTRACT: Aminoimidazole ribonucleotide (AIR) synthetase (PurM) catalyzes the conversion of formylglycinamide ribonucleotide (FGAM) and ATP to AIR, ADP, and Pi, the fifth step in de novo purine biosynthesis. The ATP binding domain of the E. coli enzyme has been investigated using the affinity label [14C]-p-fluorosulfonylbenzoyl adenosine (FSBA). This compound results in time-dependent inactivation of the enzyme which is accelerated by the presence of FGAM, and gives a Ki ) 25 µM and a kinact ) 5.6 × 10-2 min-1. The inactivation is inhibited by ADP and is stoichiometric with respect to AIR synthetase. After trypsin digestion of the labeled enzyme, a single labeled peptide has been isolated, I-X-G-V-V-K, where X is Lys27 modified by FSBA. Site-directed mutants of AIR synthetase were prepared in which this Lys27 was replaced with a Gln, a Leu, and an Arg and the kinetic parameters of the mutant proteins were measured. All three mutants gave kcats similar to the wild-type enzyme and Kms for ATP less than that determined for the wild-type enzyme. Efforts to inactivate the chicken liver trifunctional AIR synthetase with FSBA were unsuccessful, despite the presence of a Lys27 equivalent. The role of Lys27 in ATP binding appears to be associated with the methylene linker rather than its -amino group. The specific labeling of the active site by FSBA has helped to define the active site in the recently determined structure of AIR synthetase [Li, C., Kappock, T. J., Stubbe, J., Weaver, T. M., and Ealick, S. E. (1999) Structure (in press)], and suggests additional flexibility in the ATP binding region.

Aminoimidazole ribonucleotide (AIR)1 synthetase (PurM) catalyzes the fifth step of de novo purine biosynthesis: conversion of ATP and formylglycinamide ribonucleotide (FGAM) to AIR, ADP, and Pi (eq 1). The mechanism has

(1)

been proposed to involve an O-phosphorylated amide intermediate, which after attack by the N1 of the amidine of FGAM forms a five-membered ring that can lose both † This work was supported by NIH Grant GM32191. T.J.K. was supported by NIH Cancer Training Grant CA09112. * Authors to whom correspondence should be addressed. E-mail: [email protected] or [email protected]. ‡ Massachusetts Institute Technology. § Cornell University. 1 Abbreviations: AIR, aminoimidazole ribonucleotide; FGAM, formylglycinamide ribonucleotide; AIR synthetase, PurM; FGAM synthetase, PurL; FSBA, fluorosulfonylbenzoyl adenosine; LDH, lactate dehydrogenase; PK, pyruvate kinase; PEP, phosphoenolpyruvate; NADH, reduced β-nicotinamide adenine dinucleotide phosphate; BSA, bovine serum albumin; ATP, adenosine 5′-triphosphate; DTT, dithiothreitol; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; PTH, phenylthiohydantoin; TAE, Tris-acetate EDTA; TBE, Tris-borate EDTA; TFA, trifluoroacetic acid; TEA‚OAc, triethylammonium acetate; IPTG, isopropyl β-D-thiogalactopyranoside; ss DNA, single-stranded DNA; ds DNA, double-stranded DNA; wt, wild-type; nd, not determined.

phosphate and a proton to generate AIR (1). A similar mechanism, using ATP to enhance the dehydration of amides, has been proposed for FGAM synthetase (PurL) (2), the fourth step in purine biosynthesis, and for cytidine triphosphate (CTP) synthetase, a key enzyme in pyrimidine biosynthesis (3). The structure of the Escherichia coli AIR synthetase in the absence of ligands has recently been solved (PDB id 1CLI) and represents a new class of ATP binding protein (4). Our recent sequence alignments between PurMs and a subclass of the PurLs suggest that they may belong to a new superfamily of ATP-requiring enzymes catalyzing similar chemical reactions. This paper describes our efforts, using affinity labeling and site-directed mutagenesis, to define the ATP binding domain of E. coli AIR synthetase. Identification of a single labeled peptide provides a starting point for our interpretation of our structural data. A number of ATP affinity labels have been synthesized and used to identify residues within ATP binding domains (5, 6). Fluorosulfonylbenzoyl adenosine (FSBA, Chart 1) can modify several nucleophilic amino acids, yielding protein adducts of differing chemical stability. However, only in the case of Lys (7) and Tyr (8) have the adducts of FSBA been sufficiently stable to allow isolation and identification of modified peptides. Other nucleophilic amino acid residues (Arg, His, Cys, and Ser) have been implicated in reaction with FSBA based on chemical reactivity studies, but the adducts have not been isolable (9-12). FSBA has thus proven itself to be a useful reagent in the identification of a variety of amino acids implicated in ATP binding.

10.1021/bi990638r CCC: $18.00 © 1999 American Chemical Society Published on Web 07/13/1999

9832 Biochemistry, Vol. 38, No. 31, 1999 Chart 1

Labeling of specific amino acids with an affinity label is not always an indication that these residues are required for binding or catalysis. Numerous examples show the importance of confirming the proposed role of amino acids through the use of a complementary methodology such as sitedirected mutagenesis (13, 14) or structure determination (15, 16). This paper describes the use of FSBA (17) as an ATP affinity label for E. coli AIR synthetase (1). Enzyme inactivation is shown to occur stoichiometrically with specific labeling of a single amino acid residue. The peptide containing this residue has been isolated and sequenced, identifying the modified residue as Lys27. Site-directed mutagenesis of this residue has provided further insight into the ATP binding site. A comparison between the E. coli AIR synthetase and the trifunctional chicken liver protein containing AIR synthetase using the same approach is described (18, 19). This work is complementary to the recent structural determination of AIR synthetase in the absence of any organic ligands (4). MATERIALS AND METHODS Materials. Sephadex G-25 and G-50 were obtained from Pharmacia. E. coli lactate dehydrogenase (LDH, 860 units/ mg), pyruvate kinase (PK, 470 units/mg), phosphoenolpyruvate (PEP), reduced β-nicotinamide adenine dinucleotide phosphate (NADH), gelatin, bovine serum albumin (fraction V, BSA), FSBA, and adenosine 5′-triphosphate (ATP) were purchased from Sigma Chemical Co. [γ-32P]-ATP (6000 Ci/ mmol) and [8-14C]FSBA (55 mCi/mmol) were obtained from New England Nuclear. HPLC reverse-phase C18 columns (4.6 × 250 mm) used in purification of peptides were purchased from Vydac. Dithiothreitol (DTT) was purchased from Amresco (Solon, OH). S¢intA scintillation fluid was purchased from Packard. The synthesis of FGAM, as a 40: 60 R/β mixture, was accomplished as described by Mueller (20). FGAM is a mixture of anomers, and concentrations are given for the β anomer. Taq polymerase was from PerkinElmer. T4 DNA ligase was purchased from BRL. All restriction enzymes as well as HindIII- and BstEII-digested λ DNA molecular weight standards were purchased from New England Biolabs. Oligonucleotides for sequencing and site-directed mutagenesis were obtained from the MIT Biopolymers Lab. Mermaid, glass beads for purification of oligomers 200 bp, was purchased from the United States Biochemical Corp. E. coli strain TX635 (F′ lacZ+ cI857), which contains an episome-borne temperature-sensitive λ repressor (21), pJS119, a plasmid containing purM under the control of a λpL promoter (22), and TX393 [ara ∆ (lac) purM srlC300:: Tn10 recA56] (23) were gifts from Dr. John Smith and

Mueller et al. Louise Thomas of Seattle Biomedical Research Institute, Seattle, WA. General Methods. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) samples were prepared and run as described by Laemmli (24) and developed using the alternate fix and staining technique described by Cooper (25). Protein assays were accomplished using the method of Lowry et al. (26) with BSA [279 ) 0.667 mL mg-1 cm-1 (27)] as a standard. Protein sequencing was effected using an Applied Biosystems Model 477 protein sequencer with an on-line Model 120 PTH amino acid analyzer by the MIT Biopolymers Lab. Small-scale (∼10 µg) plasmid preparations were accomplished using the Magic Miniprep system from Promega, and larger scale plasmid isolation was effected using the Maxiprep kit from Qiagen. The DNA of all site-directed mutants prepared in this work was sequenced using a kit from BRL Life Sciences which employs Sequenase and the dideoxy chain termination method (28). Mass spectral analysis was performed on an Applied Biosystems Biopolymer Mass Analyzer, BIO-ION 20. Radioactivity was quantitated using a Packard 1500 TriCarb Scintillation Counter. Purification of short (100 colonies were analyzed using restriction analysis and subsequent DNA sequencing of the entire purM gene. All transformants sequenced contained the expected base changes. Defining Growth Conditions for TX635/pJS119 K27X Mutants. To optimize the growth conditions for pJS119 [containing wild type (wt) and mutants] in E. coli strain TX635, the bacteria were grown in LB media at 30 °C. Aliquots were removed just before heat induction to 42 °C and then at 1, 2, 3, and 4 h after induction. Each sample was spun for 1 min in an Epifuge at 4 °C, the supernatant was removed, and the pellet was quick-frozen in liquid nitrogen. Each sample was then resuspended in 10 mM TrisHCl (pH 7.5) such that an aliquot would produce an A600 of 0.075 OD. A 40 µL aliquot of this resuspended sample was added to 10 µL of 5× Laemmli buffer (24) and placed in a boiling water bath for 5 min. These samples were then loaded onto a 10% acrylamide gel and subjected to SDS-PAGE. The doubling time of each overproducing strain was 0.9 h, identical to the pJS119 wt strain. Bacteria were induced in late log phase (∼0.6-0.8 OD600) by adding an equal volume of media at 54 °C with rapid stirring. The cultures were transferred to a 42 °C incubator and allowed to grow for another 4 h. Growth of K27X Mutants of AIR Synthetase from TX393/ pJS119. Studies in the previous section revealed that mutant AIR synthetases were expressed constitutively. TX393 strains were grown at 30 °C in LB media and were harvested 3 h after their transition into stationary phase. Cells were collected by centrifugation (10 min, 2700g at 4 °C), and stored at -80 °C. Cell paste yields from a 1 L culture were 1.5-2.8 g. Isolation and Characterization of K27X Mutants. Isolation of wt and mutant E. coli AIR synthetases was as described by Schrimsher et al. (1), except that the final size exclusion column was omitted. The apparent Km,ATP for each Lys27 mutant AIR synthetase was determined. A typical reaction mixture contained the following in a final volume of 300 µL: 50 mM HEPES (pH 7.7), 3 mM MgCl2, 970 mM KCl, 230 µM β-FGAM, 2 mM PEP, 0.2 mM NADH, 10 unit of PK, 10 units of LDH, AIR synthetase, and varied amounts of ADP (0.2-5 × Km). Data were fit to the MichaelisMenten equation using the programs of Cleland (32). RESULTS Kinetics of InactiVation of AIR Synthetase by FSBA. FSBA is an ATP analogue that has been successfully used to covalently modify the ATP binding site of a number of

Inactivation of AIR Synthetase (PurM)

Biochemistry, Vol. 38, No. 31, 1999 9835

FIGURE 3: Stoichiometry of FSBA-AIR synthetase adduct measured using [14C]FSBA. The protein concentration (Lowry) and the radioactivity (scintillation counting) were determined for each sample. Extrapolation to total inactivation suggests that 1.1 adducts per AIR synthetase completely inactivate the enzyme. Table 2: Stability of the [14C]FSBA-AIR Synthetase Adducta

FIGURE 2: (A) Time-dependent inactivation of AIR synthetase by FSBA. AIR synthetase was incubated with (O) 0, (3) 11 µM, (b) 22 µM, (]) 66 µM, and (+) 440 µM FSBA. (B) Determination of the inactivation constants using eq 3 and a plot of 1/kobs vs 1/[FSBA].

time (h)

TFA (pH 2)

TEA‚OAc (pH 6)

NaPi (pH 7)

HEPES (pH 7.5)

0 24 48 84

0.99 0.98 0.98 0.96

0.99 0.98 0.98 0.97

0.99 nd 0.83 0.92

1.0 0.88 0.97 0.92

a

proteins (5, 6). In an effort to label the active site of AIR synthetase, various concentrations of FSBA were incubated with the enzyme and saturating amounts of FGAM, and at various times aliquots were removed and assayed for enzyme activity. The results shown in Figure 2A reveal timedependent inactivation. The model describing this process is given in eq 4, where E is enzyme, I is the inhibitor, Ki is the thermodynamic dissociation constant, and k2 is the rate constant for inactivation. Ki

k2

E + I {\} E‚I 98 E-I*

(4)

A replot of the data obtained from Figure 2A using eq 3 (Materials and Methods) gives a straight line (Figure 2B) from which the kinetic constants, Ki ) 25 µM and k2 ) 5.6 × 10-2 min-1, were determined (33). This Ki for FSBA is similar to the previously reported apparent Km for ATP of 65 µM (1). The rate of inactivation is 1/40th the rate constant for turnover with the natural substrate. A repetition of these experiments in the absence of FGAM gave a 2-fold reduction in the rate of inactivation, and in the presence of ADP (100 µM) gave complete protection against inactivation (data not shown). ADP, a competitive inhibitor with respect to ATP (1), was used in these protection experiments instead of ATP to avoid turnover. The effects of FGAM and ADP both suggest that inactivation occurs through a specific interaction between the inhibitor and AIR synthetase. Stoichiometry of InactiVation by FSBA. Control experiments revealed that AIR synthetase (at concentrations >0.8 mg/mL) is stable over a 90 min incubation at 30 °C in the absence of inhibitor. Thus, these conditions were chosen to examine the stoichiometry of its inactivation by FSBA. [14C]FSBA was used to follow the stoichiometry of adduct/ enzyme formation during the course of an inactivation. The results of a typical reaction are shown in Figure 3. These results indicate that despite the slow rate of inactivation, the

Data presented as fraction of label bound.

labeling is stoichiometric with 1.1 ( 0.2 FSBA per active site, and suggest that there is a specific interaction between the inhibitor and the enzyme. However, these experiments do not distinguish between a covalently bound and a tightly bound inhibitor. Stability of FSBA. The stability of FSBA was studied, under a variety of pH conditions, to ensure that the ester linkage would be stable to the conditions required for peptide purification. The hydrolysis of FSBA to adenosine was monitored using NMR spectroscopy, by the appearance of its H8 and H2 protons (δ 8.17 and 8.07, respectively). Integration of the signal associated with the H2 proton at various times allows a kinetic analysis of the hydrolysis process. From this analysis, a half-life for FSBA at pH 7.0 can be extrapolated to be ∼2 months, and at pH 8, 13 days. No measurable decomposition of inhibitor at pH 5.5 was found after 450 h. These results agree with those previously reported by Esch and Allison (34) using adenosine deaminase to monitor adenosine formation. They found the half-life of FSBA at 23 °C (pH 8) to be 3.6 days. Stability of the E. coli AIR Synthetase-FSBA Adduct. The stability of the linkage of the AIR synthetase-FSBA adduct was also determined, to define the conditions required to purify the modified FSBA-containing peptide(s). These experiments used Centricons and size exclusion filters to separate enzyme-bound [14C]FSBA adduct from 14C-labeled small molecules. Control experiments revealed that the filters used in this study caused no quenching of 14C-labeled material and that all of the radioactivity could be recovered. Results from the stability experiments are summarized in Table 2. From these data it is clear that the AIR synthetase‚ FSBA adduct is stable in all buffers tested for at least 84 h. The stability of the adduct suggests that either a Lys or a Tyr is modified by FSBA.

9836 Biochemistry, Vol. 38, No. 31, 1999

Mueller et al. Table 3: Kinetic Parameters for K27X AIR Synthetases enzyme

Km,ATP (µM)

kcat (s-1)

V/K (s-1 µM-1 × 102)

a

73 ( 14 58 ( 4.5 46 ( 3.1 12 ( 1.1

2.3 ( 0.1 3.0 ( 0.2 2.6 ( 0.1 2.5 ( 0.1

3.5 ( 0.8 5.2 ( 0.75 5.7 ( 0.60 21.0 ( 2.9

wt K27Q K27L K27R a

FIGURE 4: Proposed structure of the FSBA-modified tryptic peptide from AIR synthetase.

Isolation of the E. coli AIR Synthetase-FSBA-Labeled Peptide. FSBA-inactivated AIR synthetase (110 nmol) was subjected to carboxymethylation and trypsin digestion, and the peptides were isolated using a Vydac C18 reverse phase column with TFA/CH3CN elution as shown in Figure 1. The fractions containing the radiolabel were pooled, lyophilized, and reanalyzed using the sodium phosphate (pH 7)/acetonitrile elution shown in Figure 1 (inset). Lyophilization of the peptide to dryness at any stage during the purification resulted in extremely poor recoveries. The peptide isolated is the major adduct of FSBA inactivation and was recovered in 14% overall yield. The peptide was sequenced using the automated Edman degradation method to give I-X-G-V-V-K. This sequence is consistent with the expected AIR synthetase tryptic fragment:

where the circumflex indicates a position predicted to be cleaved by trypsin and the sequence of interest is underlined. It is interesting to note that if the Lys at position 27 were alkylated, trypsin would no longer be expected to cleave after that residue. Thus, the sequence is consistent with the adduct stability data and the fact that trypsin was used to digest the protein. In the second Edman degradation cycle, one major and one minor PTH-amino acid peak eluted between Tyrand Pro-PTH derivatives. N-(4-Carboxybenzenesulfonyl)lysine has previously been shown to elute with a similar retention time (35). The major peak is thus likely to be the PTH derivative of the FSBA-modified Lys27. Characterization of the FSBA-AIR Synthetase Adduct. In an attempt to define the structure of the FSBA adduct on the peptide, plasma desorption mass spectrometry was performed. The resulting parent peak of 1077.3 amu (data not shown) is consistent with the expected structure (Figure 4), which has a calculated MH+ of 1077.25 amu. Incubation of Chicken LiVer AIR Synthetase with FSBA. After successfully determining the site of interaction of FSBA with E. coli AIR synthetase, we were curious to find if FSBA also modified the enzyme from chicken liver in the same manner. AIR synthetase from this source is a trifunctional protein containing an N-terminal glycinamide ribonucleotide (GAR) synthetase and a C-terminal GAR-transformylase in addition to AIR synthetase (18, 19). Sequence comparisons show that the chicken liver enzyme has 68% sequence similarity and 51% identity with the E. coli enzyme. The chicken liver enzyme was incubated in the presence of FGAM and FSBA (1 mM). No inactivation was detected after 45 min (data not shown). Mutagenesis of E. coli AIR Synthetase (pJS119). With the site of interaction between FSBA and E. coli AIR synthetase

Redetermined in these studies.

determined, site-directed mutagenesis of Lys27 to Gln, Leu, or Arg was employed to substantiate the importance of this residue suggested by the labeling studies. Mutagenesis was accomplished using the method of Taylor et al. (31). Sequencing of the -84 through 173 bp region of the newly generated purM genes indicated that each of the desired mutants had been successfully prepared. However, three cytosines were found to be absent from the -50 through -52 region in all mutant constructs. Despite this, the reading frames for translation remain intact, and the proteins are expressed at high levels. However, protein production is no longer heat-inducible as is observed with the wt construct. Growth and Induction of pJS119 K27X Mutants. The doubling times for E. coli TX635/pJS119 mutants were indistinguishable from identical constructs expressing the wt enzyme when grown in LB media. The bacteria, containing mutant or wt AIR synthetase, doubled every 0.9 h. The optimum time to harvest the mutants was found to be 1-2 h after reaching the late log phase of growth; AIR synthetase levels slowly decrease after that time (data not shown). When enzyme was expressed in host strain TX393 (purM-), the growth of the strains producing mutant enzymes was also indistinguishable from the wt-expressing system, with doubling times of 1.3 h. Isolation and Determination of Km,ATP for AIR Synthetase K27X Mutants. AIR synthetase mutants were purified by a procedure identical to that described for wt AIR synthetase (1). The crucial purification step uses a C8-linked ATP affinity column. The mutant enzymes are retained in the same way as the wt protein, suggesting similarities in their ATP binding sites. Mutant proteins were ∼80% pure, as judged by SDS-PAGE, while the wt protein was purified to homogeneity using the same procedure. The kinetic parameters for the mutants were measured using substrate concentrations that are saturating for the wt enzyme (Table 3). A complete kinetic analysis of the mutants has not yet been carried out. The specific activities of the mutant enzymes were, in contrast to expectations, all slightly higher than that observed for the wt AIR synthetase. Even more surprising was that the Km,ATP for each mutant was lower than the Km,ATP (65 µM) previously determined for the wt protein (1). The K27R mutant has an apparent Km,ATP 5-fold lower than the wt enzyme. The identification of Lys27 in the ATP binding domain using FSBA suggested to us initially that it might play an important role in charge neutralization in phosphate binding. The mutagenesis results, however, suggest an alternative role for Lys in the ATP binding domain is required. DISCUSSION FSBA interacts with E. coli AIR synthetase as an active site affinity label, covalently modifying the enzyme specifically at Lys27. ADP protects AIR synthetase against

Inactivation of AIR Synthetase (PurM)

Biochemistry, Vol. 38, No. 31, 1999 9837

FIGURE 5: Alignment of AIR synthetases from various organisms. The alignment was calculated using the full sequence of each protein, and ClustalW (36). The coil at the top indicates the kinked helix (R1). Shaded residues are absolutely conserved. The position of Lys27 is boxed. Mesophile (top) and thermophile (bottom) sequences are separated because of differences in the Gly-rich loop region. The amino terminus of each protein is shown in the figure unless the number of the first amino acid in the figure is given in brackets. Key to abbreviations: bacteria and plants: Ec, E. coli (23); St, S. typhimurium (J. L. Zillies and D. M. Downs, unpublished observations); Hi, H. influenzae (38); At, A. thaliana [59] (39); Vu, V. unguiculata [61] (P. M. C. Smith, A. J. Mann, D. J. Hall, and C. A. Atkins, unpublished observations); Bs, B. subtilis (40); Ll, L. lactis (C. Coward, unpublished observations); Sy, Synechocystis (41); flies: Dp, D. pseudoobscura [435] (42); Dm, D. melanogaster [435] (42); Ct, C. tentans [437] (43); vertebrates: Mm, M. musculus [429] (44); Hs, H. sapiens [429] (45); Gg, G. gallus [429] (45); yeasts: Yl, Y. lipolytica [432] (C. A. Strick, L. C. James, K. E. Cole, and L. A. Elsenboss, unpublished observations); Sc, S. cereVisiae [443] (46); Sp, S. pombe [430] (47); mycobacteria: Ml, M. leprae [18] (D. R. Smith and K. Robison, unpublished observations); Mt, M. tuberculosis (48); thermophilic bacteria: Aa, A. aeolicus (49); Ph, P. horikoshii (50); thermophilic archaebacteria: Af, A. fulgidus (51); Mj, M. janaschii; Mh, M. thermoautotrophicum (52).

inhibition, while addition of FGAM to the inactivation mixture enhances the rate of inactivation. These studies suggest that FSBA acts on E. coli AIR synthetase at or near the ATP binding site. The N-terminal sequences of 24 AIR synthetase genes from different organisms are aligned in Figure 5 using ClustalW (36). Although these enzymes share a large number of conserved residues, the Lys labeled by FSBA in the E. coli enzyme is not absolutely conserved. Human AIR synthetase has a Gln residue, and Arabidopsis AIR synthetase contains an Ala in this position. Furthermore, the chicken liver trifunctional enzyme, even though it contains a Lys at this position, is not inactivated by FSBA in the presence or absence of FGAM. The lack of complete conservation of Lys27 between organisms and the fact that the trifunctional protein is not inactivated foreshadow the results of the sitedirected mutagenesis studies and provide insight about the function of this residue. To further explore the function of Lys27 in E. coli AIR synthetase, it was mutated to Gln, Leu, and Arg. The apparent Km,ATP for each mutant was determined at 230 µM FGAM. Observation of Km,ATP values for the Leu and Gln mutants that are almost identical to that for the wt enzyme establishes that Lys27 clearly is not involved in charge neutralization of the phosphates of ATP. All of the mutant residues have hydrophobic side chains that these results suggest are important for ATP binding. This contribution may be direct or conformational. Even more striking is that the turnover number for the Gln mutant is 30% greater than that for the wt protein. The basis for the improved efficiency of mutant AIR synthetase over wt is at present unclear.

We have recently determined the crystal structure of sulfate-liganded E. coli AIR synthetase modified with an N-terminal hexa-His tag at 2.5 Å resolution (4). The structure reveals a dimer held together by extensive intersubunit interactions, in agreement with solution studies on this protein (1). The interfaces between the subunits create two large clefts, each of which is lined with many conserved residues (Figure 6). A sulfate ion is bound at the wider end of each cleft, and a flexible Gly-rich hairpin loop is 25 Å away at the other end of each cleft. A kinked helix extends in an arch over the cleft, presumed to be the active site. This helix separates the Gly-rich loop from a globular central core composed of domains from each subunit. The kinked helix (residues 17-31) is visible along its entire length in only one of the subunits. In the second subunit, residues 5 through 20 are not detectable. Lys27 is located at the kink in the helix (Figure 6). The bound sulfate could be indicative of a phosphate binding site. However, it is unclear whether such a site would be associated with FGAM binding, ATP binding, or both. The FSBA inactivation studies provide important information allowing formulation of a model for substrate binding within the active site. FSBA acylates Lys27 in the kinked helix (Figure 6). These results have been interpreted to indicate that ATP binds near the Gly-rich hairpin, 20-30 Å away from the sulfate binding site. If our interpretation of the FSBA binding studies is correct, then the sulfate is probably indicative of the FGAM phosphate binding site. Sequence conservation, sulfate binding, and FSBA labeling together indicate that the cleft defines the active site of AIR synthetase. The FSBA studies have been particularly critical

9838 Biochemistry, Vol. 38, No. 31, 1999

FIGURE 6: Stereoview of the putative active site cleft in E. coli AIR synthetase and the position of the sulfate relative to Lys27. A molecular surface is shown for the AIR synthetase dimer at the subunit interface, which defines the active site cleft. Green surfaces correspond to solvent-exposed, strictly conserved residues. Important structural features discussed in the text are labeled. The white ribbon is the kinked helix R1 (residues 5-34) from the lighter colored subunit 1. Sulfate and Lys27, which is located at the kink in the helix, are presented in CPK space-filling models with traditional colors.

as thus far we have been unable to crystallize either ATP or an FGAM analogue with AIR synthetase. Lys residues are typical components of ATP binding sites, where they frequently interact electrostatically with the phosphate moieties. Several observations indicate that Lys27 in E. coli AIR synthetase may not perform this function and that it is located in a conformationally dynamic region of the enzyme. The crystal structure reveals that the side chain of Lys27 including its amino group points away from the putative active site cleft. This conformation leaves only the β- and γ-methylenes of Lys adjacent to the putative ATP binding pocket (Figure 6). Our mutagenesis studies indicate that the methylene linker and not the amino group of Lys plays a key role in defining the ATP binding domain. The unliganded structure and the positioning of Lys27 are probably not indicative of the active protein conformation. As noted above, Lys27 is located on a dynamic kinked helix adjacent to a glycine-rich flexible loop (Figure 6). The flexibility of the helix is apparent by the differences in the N-terminus of the two subunits of AIR synthetase (4). In addition, the flexibility of the glycine-rich loop is indicated by this loop’s higher than average B factors. Finally, efforts to soak ATP into apo AIR synthetase crystals resulted in their cracking. Thus, the structure of apo AIR synthetase is probably not indicative of the conformation of Lys27 with the FSBA bound. The efficiency of FSBA modification, its potentiation by FGAM, and its inhibition by ADP support labeling of the ATP binding domain as well as its conformational flexibility. The distance between Lys27 and the sulfate (Figure 6) suggests that an extended or partially extended nucleotide and FGAM can fit into the putative active site cleft. Definition of the detailed binding of ATP and the role of Lys27 in the active site will require a structure determination in the presence of nucleotides. Subtle differences in conformation may also account for the observation that the trifunctional AIR synthetase is not

Mueller et al. inactivated by FSBA. This result was unexpected as the proteins are homologous and both contain Lys27. A major difference between the mono- and trifunctional synthetases, however, is that the trifunctional protein is tethered to GAR synthetase at its N-terminus. This fusion could affect the conformation of the flexible kinked helix and its short extension (16 residues to the N-terminus). Thus, despite the fact that Lys27 is present in the trifunctional protein, the additional protein domain attached to the N terminus of the trifunctional protein could affect lid movement over the active site cleft in a fashion different from the monofunctional AIR synthetase. There is substantial literature (11, 37) in which FSBA and other putative ATP analogues have provided information about the region surrounding the ATP binding site of different proteins. The tight binding of FSBA to E. coli AIR synthetase favors the hypothesis that FSBA has labeled a residue in the region in which ATP binds. Further biochemical studies in conjunction with structural information in the presence of nucleotides are required for further definition of the role of residue 27 in ATP binding. REFERENCES 1. Schrimsher, J. L., Schendel, F. J., Stubbe, J., and Smith, J. M. (1986) Biochemistry 25, 4366-4371. 2. Schendel, F. J., Mueller, E., Stubbe, J., Shiau, A., and Smith, J. M. (1989) Biochemistry 28, 2459-2471. 3. Hardie, D. G., and Coggins, J. R. (1986) in Multidomain Proteins: Structure and EVolution (Hardie, D. G., and Coggins, J. R., Eds.) pp 1-12, Elsevier, New York. 4. Li, C., Kappock, T. J., Stubbe, J., Weaver, T. M., and Ealick, S. E. (1999) Structure (in press). 5. Colman, R. F. (1983) Annu. ReV. Biochem. 52, 67-91. 6. Colman, R. F. (1989) in Protein Function, A Practical Approach (Creighton, T. E., Ed.) pp 77-99, IRL Press, Oxford. 7. Pinkofsky, H. B., Ginsburg, A., Reardon, I., and Heinrikson, R. L. (1984) J. Biol. Chem. 259, 9616-9622. 8. Bitar, K. G. (1982) Biochem. Biophys. Res. Commun. 109, 30-35. 9. Poulos, T. L., and Price, P. A. (1974) J. Biol. Chem. 249, 1453-1457. 10. Krieger, T. J., and Miziorko, H. M. (1986) Biochemistry 25, 3496-3501. 11. Pandey, V. N., and Modak, M. J. (1988) J. Biol. Chem. 263, 6068-6073. 12. Harlow, K. W., and Switzer, R. L. (1990) J. Biol. Chem. 265, 5487-5493. 13. Parakh, C. R., and Villafranca, J. J. (1990) Abstr. Pap.sAm. Chem. Soc. 199, BIOL-111. 14. Freitag, N. E., and McEntee, K. (1991) J. Biol. Chem. 266, 7058-7066. 15. Vollmer, S. H., and Colman, R. F. (1990) Biochemistry 29, 2495-2501. 16. Kohlstaedt, L. A., Wang, J., Friedman, J. M., Rice, P. A., and Steitz, T. A. (1992) Science 256, 1783-1790. 17. Pal, P. K., Wechter, W. J., and Colman, R. F. (1975) J. Biol. Chem. 250, 8140-8147. 18. Daubner, S. C., Schrimsher, J. L., Schendel, F. J., Young, M., Henikoff, S., Patterson, D., Stubbe, J., and Benkovic, S. J. (1985) Biochemistry 24, 7059-7062. 19. Schrimsher, J. L., Schendel, F. J., and Stubbe, J. (1986) Biochemistry 25, 4356-4365. 20. Mueller, E. J. (1994) Ph.D. Thesis, Massachusetts Institute of Technology. 21. Mieschendahl, M., and Mu¨ller-Hill, B. (1985) J. Bacteriol. 164, 1366-1369. 22. Inglese, J., Johnson, D. L., Shiau, A., Smith, J. M., and Benkovic, S. J. (1990) Biochemistry 29, 1436-1443.

Inactivation of AIR Synthetase (PurM) 23. Smith, J. M., and Daum, H. A., III (1986) J. Biol. Chem. 261, 10632-10636. 24. Laemmli, U. K. (1970) Nature 227, 680-685. 25. Cooper, T. (1977) The Tools of Biochemistry, John Wiley & Sons, New York. 26. Lowry, O. H., Rosenbrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. 27. Schachman, H. D., and Edelstein, S. J. (1966) Biochemistry 5, 2681-2675. 28. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 5463-5467. 29. Sambrook, J., Fritsch, G. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Guide, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 30. Foster, W. B., Griffith, M. J., and Kingdon, H. S. (1981) J. Biol. Chem. 256, 882-886. 31. Taylor, J. W., Ott, J., and Eckstein, F. (1985) Nucleic Acids Res. 13, 8765-8785. 32. Cleland, W. W. (1979) Methods Enzymol. 63, 103-138. 33. Main, A. R. (1973) in Essays in Toxicology (Hays, J. W., Eds) pp 59-105, Academic Press, New York. 34. Esch, F. S., and Allison, W. S. (1978) J. Biol. Chem. 253, 6100-6106. 35. Saradambal, K. V., Bednar, R. A., and Colman, R. F. (1981) J. Biol. Chem. 256, 11866-11872. 36. Thompson, J. D., Higgins, D. G., and Gibson, T. J. (1994) Nucleic Acids Res. 22, 4673-4680. 37. Beese, L. S., Friedman, J. M., and Steitz, T. A. (1993) Biochemistry 32, 14095-14101. 38. Fleischmann, R. D., Adams, M. D., White, O., Clayton, R. A., Kirkness, E. F., Kerlavage, A. R., Bult, C. J., Tomb, J. F., Dougherty, B. A., Merrick, J. M., McKenney, K., Sutton, G., Fitzhugh, W., Fields, C., Gocayne, J. D., Scott, J., Shirley, R., Liu, L. I., Glodek, A., Kelley, J. M., Weidman, J. F., Phillips, C. A., Spriggs, T., Hedblom, E., Cotton, M. D., Utterback, T. R., Hanna, M. C., Nguyen, D. T., Saudek, D. M., Brandon, R. C., Fine, L. D., Fritchman, J. L., Fuhrmann, J. L., Geoghagen, N. S. M., Gnehm, C. L., McDonald, L. A., Small, K. V., Fraser, C. M., Smith, H. O., and Venter, J. C. (1995) Science 269, 496-512. 39. Senecoff, J. F., and Meagher, R. B. (1993) Plant Physiol. 102, 387-399. 40. Ebbole, D. J., and Zalkin, H. (1987) J. Biol. Chem. 262, 82748287. 41. Kaneko, T., Tanaka, A., Sato, S., Kotani, H., Sazuka, T., Miyajima, N., Sugiura, M., and Tabata, S. (1995) DNA Res. 2, 153-166. 42. Henikoff, S., and Eghtedarzadeh, M. K. (1987) Genetics 117, 711-725. 43. Clark, D. V., and Henikoff, S. (1992) J. Mol. EVol. 35, 5159. 44. Kan, J. L., Jannatipour, M., Taylor, S. M., and Moran, R. G. (1993) Gene 137, 195-202.

Biochemistry, Vol. 38, No. 31, 1999 9839 45. Aimi, J., Qiu, H., Williams, J., Zalkin, H., and Dixon, J. E. (1990) Nucleic Acids Res. 18, 6665-6672. 46. Henikoff, S. (1986) J. Mol. Biol. 190, 519-528. 47. McKenzie, R., Schuchert, P., and Kilbey, B. (1987) Curr. Genet. 12, 591-597. 48. Cole, S. T., Brosch, R., Parkhill, J., Garnier, T., Churcher, C., Harris, D., Gordon, S. V., Eiglmeier, K., Gas, S., Barry, C. E., III, Tekaia, F., Badcock, K., Basham, D., Brown, D., Chillingworth, T., Connor, R., Davies, R., Devlin, K., Feltwell, T., Gentles, S., Hamlin, N., Holroyd, S., Hornsby, T., Jagels, K., Krogh, A., McLean, J., Moule, S., Murphy, L., Oliver, K., Osborne, J., Quail, M. A., Rajandream, M. A., Rogers, J., Rutter, S., Seeger, K., Skelton, J., Squares, R., Squares, S., Sulston, J. E., Taylor, K., Whitehead, S., and Barrell, B. G. (1998) Nature 393, 537-544. 49. Deckert, G., Warren, P. V., Gaasterland, T., Young, W. G., Lenox, A. L., Graham, D. E., Overbeek, R., Snead, M. A., Keller, M., Aujay, M., Huber, R., Feldman, R. A., Short, J. M., Olsen, G. J., and Swanson, R. V. (1998) Nature 392, 353358. 50. Kawarabayasi, Y., Sawada, M., Horikawa, H., Haikawa, Y., Hino, Y., Yamamoto, S., Sekine, M., Baba, S., Kosugi, H., Hosoyama, A., Nagai, Y., Sakai, M., Ogura, K., Otsuka, R., Nakazawa, H., Takamiya, M., Ohfuku, Y., Funahashi, T., Tanaka, T., Kudoh, Y., Yamazaki, J., Kushida, N., Oguchi, A., Aoki, K., and Kikuchi, H. (1998) DNA Res. 5, 147-155. 51. Klenk, H. P., Clayton, R. A., Tomb, J. F., White, O., Nelson, K. E., Ketchum, K. A., Dodson, R. J., Gwinn, M., Hickey, E. K., Peterson, J. D., Richardson, D. L., Kerlavage, A. R., Graham, D. E., Kyrpides, N. C., Fleischmann, R. D., Quackenbush, J., Lee, N. H., Sutton, G. G., Gill, S., Kirkness, E. F., Dougherty, B. A., McKenney, K., Adams, M. D., Loftus, B., Peterson, S., Reich, C. I., McNeil, L. K., Badger, J. H., Glodek, A., Zhou, L. X., Overbeek, R., Gocayne, J. D., Weidman, J. F., McDonald, L., Utterback, T., Cotton, M. D., Spriggs, T., Artiach, P., Kaine, B. P., Sykes, S. M., Sadow, P. W., DAndrea, K. P., Bowman, C., Fujii, C., Garland, S. A., Mason, T. M., Olsen, G. J., Fraser, C. M., Smith, H. O., Woese, C. R., and Venter, J. C. (1997) Nature 390, 364-370. 52. Smith, D. R., Doucette-Stamm, L. A., Deloughery, C., Lee, H., Dubois, J., Aldredge, T., Bashirzadeh, R., Blakely, D., Cook, R., Gilbert, K., Harrison, D., Hoang, L., Keagle, P., Lumm, W., Pothier, B., Qiu, D., Spadafora, R., Vicaire, R., Wang, Y., Wierzbowski, J., Gibson, R., Jiwani, N., Caruso, A., Bush, D., Safer, H., Patwell, D., Prabhakar, S., McDougall, S., Shimer, G., Goyal, A., Pietrokovski, S., Church, G. M., Daniels, C. J., Mao, J. I., Rice, P., Nolling, J., and Reeve, J. N. (1997) J. Bacteriol. 179, 7135-7155. BI990638R